THE EARTHQUAKE ADVANCE THEORY 197 probable than we do now; but if their fronts are pushed forward we should be inclined to consider this the most rational of the hypotheses proposed. It is important to note that the former great advance of Hidden and Nunatak Glaciers 15 to 20 miles beyond their present termini is not so remarkable as may seem at first thought. If, for instance, Hubbard and Turner Glaciers should be pushed out two or three miles, which, in view of the recent advance of the much smaller and less active Hidden Glacier, is certainly a conceivable result of advance due to earthquake .shaking) the entire Russell Fiord and Nunatak Fiord would be transformed to a lake by an ice dam at the head of Disenchantment Bay. The tideless, fresh water of this lake would cause less recession in Nunatak Glacier than the tidal salt water now does. Mpreover, the lake water would be much colder than the waters of the fiord, which are influenced by direct connection with the ocean. This coldness would be increased by the icebergs from Nunatak Glacier, which could no longer escape to the Pacific, and by those from Hidden, Variegated, Orange, and perhaps smaller glaciers which would become tidal. It is even conceivable that the lake waters would be so filled with floating ice that they would remain frozen well into the summer. The effect of all these conditions would be to aid in the extension of Nunatak and Hidden Glaciers in two very important ways: (1) by greatly checking the discharge of icebergs from the ice fronts, (2) by lowering the mean annual temperature of the region near by, and thus diminishing ablation and increasing snowfall. Thus a thrust which pushed Hubbard and Nunatak Glaciers forward a distance of only a few miles might result in a far greater extension of Hidden and Nunatak Glaciers. A part of this extension would need no other impelling cause than that which now exists; for with lower mean annual temperature, and lessened iceberg discharge the fronts of the glaciers would necessarily move forward beyond the position which they now hold as a balance between supply and wastage from ablation and iceberg discharge. When the ice dam across Russell Fiord was removed by recession of Hubbard Glacier, the return of the tidal water to the margin of the advanced Nunatak Glacier would inaugurate recession at a more rapid rate, and if the cause for the advance had by that time died out, the glaciers might be expected to recede rapidly. These considerations are of course as applicable to an advance by other causes as to an advance in response to earthquake shaking. But since these associated phenomena make it clear that so great an advance as that of Nunatak and Hidden Glaciers does not necessarily mean an actual forward thrust of a glacier for 15 or 20 miles by the sole and unaided cause of response to earthquake shaking, they remove some of the objection to this hypothesis on the score of the enormous forward movement which actually occurred. Only a part of the advance is assigned by this hypothesis to the direct thrust; the remainder is ascribed to the result of accessory conditions indirectly the result of advance under the impulse of earthquake shaking. With this advance, and until further evidence is available, we believe that the earthquake hypothesis has basis for retention as a possible explanation of the remarkable advance of the Yakutat Bay glaciers, from which these glaciers were withdrawing when the earthquakes of 1899 gave rise to conditions by which new advance was necessitated in some of them, and perhaps, ultimately, in all the more important ones.